Low offset op amp for audio

[saint gillis]

Hi, the subject has of course been talked here and here on the forum, but let's be more specific : what are the op amp candidates today that combine low distortion, low noise, and low enough DC offset to avoid AC coupling in a design ? And let's add another constraint : a less than 3 or 5€ cost

 

[JohnRoberts]

Be aware that there are chopper stabilized op amps that delivered extremely low DC offsets (the DC chopper alternates inputs to cancel DC errors). These can deliver spectacular DC characteristics but with the cost of a little extra switching noise.
==

Check the op amp data sheets. Most DC coupled audio paths that I've seen use some form of DC servo to subtract out DC errors.

Good luck

JR

 

[saint gillis]

I have used successfully an OPA2134 on a small section of a circuit where I wanted to avoid to AC couple and it worked, even if the noise specs are good, it seems we can find better nowadays.
The OPA2156 seems promising on the paper..

 

[saint gillis]

ADA4511-2
ADA4522-1

 

[cyrano]

ADA4511-2

Pretty impressive. Up to +/- 45 V supply. Under 3 mA current needed. I'm a bit of an Analog Devices fanboy, I'm afraid...

Even at the "normal" +/- 20 V this should provide some swing!

 

[jensenmann]

OPA2132 was the DC precision version of the OPA2134. They became prohibitively expensive in the last few years, though....

 

[FETlife]

Hi, the subject has of course been talked here and here on the forum, but let's be more specific : what are the op amp candidates today that combine low distortion, low noise, and low enough DC offset to avoid AC coupling in a design ? And let's add another constraint : a less than 3 or 5€ cost

How are you defining 'low enough DC offset'? Does OPA167x suit?

 

[saint gillis]

The TL07x is 1mV typical, 4mV max, I've seen designs with potentiometers DC coupled to TL07Xs working for a while and becoming scratchy after few years, so I'd say 100µV typical would be the maximum tolerable

 

[Voyager10]

What is the application? How much gain is involved?

 

[dom]

Have you considered using a servo?

 

[Newmarket]

The TL07x is 1mV typical, 4mV max, I've seen designs with potentiometers DC coupled to TL07Xs working for a while and becoming scratchy after few years, so I'd say 100µV typical would be the maximum tolerable

Isn't that scratchy Ness a function of potentiometer track/wiper wear'n'tear though ? Rather than due to the op amp output DC offset.

 

[saint gillis]

Isn't that scratchy Ness a function of potentiometer track/wiper wear'n'tear though ? Rather than due to the op amp output DC offset.

Maybe

 

[saint gillis]

What is the application? How much gain is involved?

I have no specific application in mind, but you are right, it must be considered

 

[JohnRoberts]

Isn't that scratchy Ness a function of potentiometer track/wiper wear'n'tear though ? Rather than due to the op amp output DC offset.

I haven't seen pot's screened resistive substrates literally worn away, but I guess in theory it is possible. I don't deal with much really old gear. I have seen dirt and contaminants interfere with smooth continuous wiper electrical contact.

The mechanism for DC voltage across a pot substrate causing a scratchy noise is due to tiny voltage steps or jumps as the wiper sweeps across the pot element. Pot wipers typically use multiple fingers (5 is a common number) to improve contact during wiper bounce.

JR

 

[Newmarket]

I haven't seen pot's screened resistive substrates literally worn away, but I guess in theory it is possible. I don't deal with much really old gear. I have seen dirt and contaminants interfere with smooth continuous wiper electrical contact.

The mechanism for DC voltage across a pot substrate causing a scratchy noise is due to tiny voltage steps or jumps as the wiper sweeps across the pot element. Pot wipers typically use multiple fingers (5 is a common number) to improve contact during wiper bounce.

JR

Oh it's definitely possible. And I have personally been the cause of ruined CP pot tracks and the diagnosis of customer issues with this when working at P&G in South Wales
Will post more when on a proper keyboard/ PC.

 

[JohnRoberts]

I didn't say it wasn't possible just that I never encountered it.

I rarely used P&G faders, they were pretty expensive IIRC. We used them for one customer back in the 70s/80s at Loft.

JR

PS At Peavey $5 was an expensive fader. I tried to spec a $10 Alps fader into my big AMR 36x24 split console and my boss (the guy with his name on the buildings) veto'd my choice. He was actually cheaper than me, if that is possible. Console customers feel the faders before listening to a console.

 

[Newmarket]

I didn't say it wasn't possible just that I never encountered it.

Yes. That's what I understood you to mean.
It's rare with decent quality CP tracks. I did it by means of a poorly conceived microcontroller based moving fader system And I did get some returns to look at where there was excessive track wear due to a customer's own over eager moving fader servo.
On the tracks themselves - there was some belief / myth that the 'ink' used on older faders was better. But the manufacturer stopped making it and so the ink changed.

I rarely used P&G faders, they were pretty expensive IIRC. We used them for one customer back in the 70s/80s at Loft.

Expensive. Yes. Sort of a hierarchy of customers - SSL etc and "Mission Critical" broadcast customers were P&G by default (though IIRC SSL started to move away from that for some product). "Mid-Range" eg Soundcraft / Soundtracs offered P&G as an option putting maybe +£2K on the end cost of 48 Channels.
Alps and TKD the major competitors there.
8000 series (carbon loaded nylon bodies) was aimed as lower cost option relative to 3000 Series.

JR

PS At Peavey $5 was an expensive fader. I tried to spec a $10 Alps fader into my big AMR 36x24 split console and my boss (the guy with his name on the buildings) veto'd my choice. He was actually cheaper than me, if that is possible. Console customers feel the faders before listening to a console.

Yes. And first thing they do at trade shows. And the 'feel' can be adjusted to suit the customer - not really related to audio quality but often perceived to be so.

 

[JohnRoberts]

We were a pretty good customer for Alps and I wanted to use their $10 fader for my big split console but I was over ruled, and we switched to a new fader from Korea made by Jung Poon. It used the two guide rods internal configuration to stabilize the slider assembly and "felt" like a more expensive fader. That was until the plastic wiper assembly broke, and the fader stopped working. They were set to win huge market share in the mixer business except for that one design/production flaw.

Coincidentally Alps introduced their "K fader" (using two guide rods) in response for $5 and won back the business. I was still sitting on several thousand of the dodgy Jung Poon faders. I recall a rather uncomfortable meeting between me, the Jung Poon US distributer, and Peavey's director of purchasing. My purchasing director hung me out to dry and wasn't very supportive, but I got the distributer to agree that my QC screening test of pushing the wiper blade from side to side was valid. He accepted back a couple thousand reject faders that day, that I had culled out by testing every fader we had in inventory.

JR

 

[Newmarket]

Yeah...Quality Issues. Long time back now but I recall someone at AMEK "congratulating" P&G on their Quality "Control" as they had obviously stopped it getting out of the factory

 

[JohnRoberts]

Come to think of it, Alps has a pretty good quality reputation while engineered for a lower price point.

I had one custom rotary pot tooled up for the EQ in that same split console that used a special dual track screened resistance element that was shorted across by the wiper fingers. We sourced two vendors. The US vendor just used a standard 5 finger wiper, Alps tooled up a special split wiper that made better track contact and suffered noticeably less audible wiper bounce. That EQ circuit was unstable during wiper bounce so sounded like a scratchy pot but it was actually momentary bursts of instability. The Alps part was superior between those two.

JR